Based on the climate-driven yield index (CDYI) series of summer maize in 17 cities and the monthly series of large-scale atmospheric circulation indices (LACI), the responses of yield fluctuations of summer maize to atmospheric circulation anomalies during 1988–2017 were explored in Henan province, north China. The main findings were as follows: (1) with using principal component analysis, east, north, central, and west Henan were identified as four sub-regions with different CDYI variations; (2) with using ensemble empirical mode decomposition, the CDYI in east and north Henan presented a notable 6-year oscillation, while central and west Henan were dominated by a 3-year oscillation; (3) East Pacific/North Pacific Oscillation in January, Eastern Tropical Pacific SST in May, Tropical Southern Atlantic Index in June, and North Pacific in September were the optimal yield prediction signals in east, north, central, and west Henan, respectively; (4) the regression model with predictors as year and various LACI had the ideal forecasting results on actual regional maize yield, with the average relative error ranged from 4.94 to 9.20%.